Project

General

Profile

Actions

Bug #4491

closed

mds: assert failure on _purge_forward_pointers

Added by Sam Lang about 11 years ago. Updated about 11 years ago.

Status:
Resolved
Priority:
Urgent
Assignee:
Category:
-
Target version:
-
% Done:

0%

Source:
Development
Tags:
Backport:
Regression:
Severity:
3 - minor
Reviewed:
Affected Versions:
ceph-qa-suite:
Component(FS):
Labels (FS):
Pull request ID:
Crash signature (v1):
Crash signature (v2):

Description

Joe Buck reported a bug with master:

INFO:teuthology.task.ceph.mds.0.err:mds/MDCache.cc: In function 'void MDCache::_purge_forwarding_pointers(inode_backtrace_t*, CDentry*, int, Context*)' thread 7f7c40c78700 time 2013-03-18 14:43:33.346322
INFO:teuthology.task.ceph.mds.0.err:mds/MDCache.cc: 8452: FAILED assert(r 0 || r -2)
INFO:teuthology.task.ceph.mds.0.err: ceph version 0.58-733-g9010b10 (9010b10c5486f0c7f57bc14b0b0bd088d0f55d27)
INFO:teuthology.task.ceph.mds.0.err: 1: (MDCache::_purge_forwarding_pointers(inode_backtrace_t*, CDentry*, int, Context*)+0x702) [0x57d662]
INFO:teuthology.task.ceph.mds.0.err: 2: (Objecter::handle_osd_op_reply(MOSDOpReply*)+0xe38) [0x6f16b8]
INFO:teuthology.task.ceph.mds.0.err: 3: (MDS::handle_core_message(Message*)+0xae8) [0x4dbbc8]
INFO:teuthology.task.ceph.mds.0.err: 4: (MDS::_dispatch(Message*)+0x2f) [0x4dbd8f]
INFO:teuthology.task.ceph.mds.0.err: 5: (MDS::ms_dispatch(Message*)+0x1db) [0x4dd82b]
INFO:teuthology.task.ceph.mds.0.err: 6: (DispatchQueue::entry()+0x341) [0x81e711]
INFO:teuthology.task.ceph.mds.0.err: 7: (DispatchQueue::DispatchThread::entry()+0xd) [0x79b95d]
INFO:teuthology.task.ceph.mds.0.err: 8: (()+0x7e9a) [0x7f7c44fd0e9a]
INFO:teuthology.task.ceph.mds.0.err: 9: (clone()+0x6d) [0x7f7c437894bd]

Using the following yaml:

overrides:
ceph:
extra_packages: [libcephfs1, libcephfs-java, libcephfs-jni, ceph-test]

roles:
- [mon.0, mds.0]
- [mon.1, osd.0, hadoop.master.0]
- [mon.2, osd.1, client.0, hadoop.slave.0]

tasks:
- ssh_keys:
- install:
- ceph:
- hadoop:
- workunit:
clients:
client.0: [hadoop-wordcount]

Actions #1

Updated by Tamilarasi muthamizhan about 11 years ago

This happens soon after ceph-fuse mount. I hit this when trying to run blogbench test.

Actions #2

Updated by Sam Lang about 11 years ago

  • Status changed from In Progress to Fix Under Review

I pushed a proposed fix to wip-4491. Basically we just need to handle the case that the osd returns ENODATA.

Actions #3

Updated by Sage Weil about 11 years ago

  • Status changed from Fix Under Review to Resolved
Actions

Also available in: Atom PDF